Technology Commons Alliance

Prometheus or Icarus

by the TCA team ·

The fire that has burned in humanity’s hearth from the beginning, the fire with which we have remade the world, is a profoundly double-edged symbol both of our Promethean power to control the earth… and of the frustratingly unexpected limits we repeatedly encounter in our exercise of that power.

William Cronon, foreword to Stephen Pyne, Fire: A Brief History (2001)

Ancient wisdom holds a curious, often contradictory relationship with fire. Even today, with our newfound particle physics and advanced understanding of states of matter, there is still an ineffable quality to the dancing lights that sit at the foundations of human culture and society.

The depth of our relationship with this phenomenon feels at times like the only universal quality we share. Yet the self-contradictory nature it holds in our histories is preserved. Contradictions of this sort, which in modern objects are seen as lines of conflict or points of division, are for fire held as one. The destructive and humbling quality described in the myth of Icarus is paired in equal part with the life-bringing and knowledge-revealing qualities that the myth of Prometheus describes as the root of human freedom: a power so consequential that its bearer had to be punished by the gods themselves.

But within both these stories, one quality remains consistent: the tantalizing, alluring dizziness with which fire captures our attention. The only line drawn between death and life is our distance from it, and how we choose to use it. In other words, these stories describe a consistent human quality.

It is not quite fair to compare AI with fire itself, but it holds the same contradictory kind of relationship. We set out to expand the limits of human possibility, to march the future into the now. It has captured our attention, but have we stopped to ask the questions our history has? Will our capacity to maintain this relationship keep pace? And what is it about us, about humans’ ability to contain not only ourselves but forces as wild and brilliant as flames, that we need to remember?

The myths we tell ourselves

We tend to read these myths as opposing accounts of human ambition: Prometheus as the champion of ingenuity, Icarus as its cautionary tale. But this distinction is perhaps too convenient. Neither story is quite so simple.

The myth of Prometheus is no inherent triumph. In the version Plato tells, Epimetheus distributes natural abilities among the animals but leaves humanity unequipped, and Prometheus steals fire and technical knowledge to compensate. The philosopher Bernard Stiegler’s reading is that we lack qualities of our own, and so are burdened, in a sense, to develop our world through tools and techniques. Technology is not simply something we invent after becoming human; it is deeply involved in what makes us human to begin with. Fire does not simply give humanity power; it makes us responsible for what we do with it.

Achieving power versus holding it

This is where our parallel with AI settles in properly. Perhaps one of the most interesting examples is in recent developments in mathematics. In September 2026, OpenAI announced an AI-assisted mathematical construction demonstrating that the Navier-Stokes equations did indeed blow up under certain conditions. These equations, which describe the movement of fluids, have challenged mathematicians for generations, and while the specifics are hotly debated, particularly the difference between forced and unforced versions of the three-dimensional Navier-Stokes equations, they remain hard enough to be considered Millennium Prize-worthy.

Whatever the technical case, it was nevertheless a remarkable demonstration of AI’s emerging ability to contribute to research at the frontier of mathematics. But it did raise questions for mathematicians about what knowledge of this kind really is. In the words of the Fields medalist Terence Tao:

Perhaps the most counter-intuitive consequence is that human understanding of mathematics is falling behind, because AI systems generate results faster than mathematicians can digest and absorb them.

Terence Tao, Nature, October 2026

While researching this, we were reminded of a line from Douglas Adams. After waiting seventy-five thousand generations, the great supercomputer Deep Thought came to the answer to life, the universe and everything:

“Forty-two!” yelled Loonquawl. “Is that all you’ve got to show for seven and a half million years’ work?”

“I checked it very thoroughly,” said the computer, “and that quite definitely is the answer. I think the problem, to be quite honest with you, is that you’ve never actually known what the question is.”

Douglas Adams, The Hitchhiker’s Guide to the Galaxy (1979), chapter 28

Adams was on to something here, all the way back in the 1970s: a particular quality of humans to take the answer for granted, missing all the content that made it meaningful.

To Tao’s point, for many years mathematical proofs were the culmination of collective, often piecemeal developments across fields; each step sparked conversations and new developments in others, until the work culminated in a written proof, a final verdict, the last page of a great novel. The value of this end point is measured in the journey it took to get there; it exemplifies that journey, but it does not contain it. What OpenAI’s demonstration really tells us is that this feeling of success is not a property of the proof itself; it is a property of our relationship with the idea of success.

The middle way

As we turn to our second myth, that of Icarus, it is easy to draw the parallel with AI as the sun. We are carried by our own hubris higher and higher until we burn ourselves. But as we said at the start, it is not quite fair to compare AI with fire itself. We think the better comparison is the wings.

The myth describes Daedalus, Icarus’s father, building them to escape his own labyrinth in Crete. The most interesting part here is that, before they took off, he gave his son a warning:

Let me warn you, Icarus, to take the middle way,
in case the moisture weighs down your wings, if you fly too low,
or if you go too high, the sun scorches them.
Travel between the extremes.

Ovid, Metamorphoses, Book VIII, lines 183–235, translated by A. S. Kline

It was not just a suggestion to avoid flying too high; not to fly high enough was death all the same. Icarus fell because flight is intoxicating. The wings made rising effortless, with the same tantalizing quality we read in fire: the ability of our tools to make real what we could never have imagined on our own.

This, we think, is where we find ourselves today. According to Stanford’s AI Index, global corporate AI investment reached nearly $582 billion in 2025, and it has not slowed in the year since. Driven by measures like computing power and benchmark performance, AI exemplifies what tools are capable of doing for us, even in its earlier stages; take, for example, DeepMind’s AlphaGo program. Our measure of AI is its ability to outperform us, and more recently, itself.

AI gives us the feeling of rushing toward the future, the proof without the years of conversation that led to it. We feel more productive, we see more results, and so we use it more, believe in it more, and fund it more. Each turn of that loop lifts us a little higher, and each one moves the answer a little further from the question.

Forty-two may well have been the answer. But an answer we cannot understand is not knowledge we hold; it is knowledge we are holding onto.

Fire before the fireplace

Daedalus also wore the same wings; he felt the same lift. But he survived because he understood what he had built. He knew how the heat would find it, and how high was too high. That, in the end, is the point. It is not enough for the builders of these systems to understand them, or for the warnings to be written somewhere in the instructions. Each of us who picks up the wings needs the capacity to evaluate them for ourselves: to know what these systems are, what they are not, and where our own limits sit.

We need to measure our own guardrails rather than wait for someone to hand them to us. This comes in the form of new measures of progress, ones that ask not how high the machine can carry us, but how we manage to “take the middle way.”

The greatest achievement of fire was never the flame itself; Prometheus gave us suffering as much as he gave us potential. Instead, the greatest achievement was the fireplace: the infrastructure that maintains our mutually beneficial relationship with it, what actually channels the potential toward human needs and human requirements.

AI in its dizzying ecstasy, its capacity to drive us upward and upward, toward goals that are measurably successful but structurally shallow, risks making us forget that the ability to harness powerful forces is what progress actually consists of. Technological advancement without corresponding human development is only half of what progress means. We have demonstrably recognized this throughout history, but seem to lose sight of it behind the veneer of measurable performance.

Daedalus’s warning was good advice, but it was not enough to save his son. Advice from the people who build our tools never will be, because what kept Daedalus alive was understanding that grew alongside the thing he made. We need our own independent capacity to understand and manage the tools that will come to define our future. Just as fire required a fireplace, we need to build a new conception of progress that can sustain our relationship with potential as much as it can build potential in the first place.

Sources

— The TCA team